It gets weirder now. I tried not using Mesos, and run Spark on standalone mode and it works.
In the process, I discovered something that might hint me...so, I basically run the spark master and let some slaves connect to it, with the following command ./bin/spark-class org.apache.spark.deploy.worker.Worker spark://srv-esx-isc01:7077 and it works fine, my slaves connect to the master. But when I try this ./bin/spark-class org.apache.spark.deploy.worker.Worker spark://10.10.0.141:7077 it does not work, and the IP address is correct... I will try changing the /etc/hosts file on every node that is supposed to run mesos and hard code the IPs cuz I am pretty sure the actor which is not found is bound to a crappy address I ll keep u posted -- View this message in context: http://apache-spark-user-list.1001560.n3.nabble.com/spark-0-9-0-on-top-of-Mesos-error-Akka-Actor-not-found-tp1164p1178.html Sent from the Apache Spark User List mailing list archive at Nabble.com.
